Elevator health monitoring system

Abstract

An elevator system includes a component adapted to perform a function, a sensor, and a control configuration. The sensor is configured to detect an operating parameter associated with the function. The control configuration is configured to receive a parameter signal from the sensor, and extract a predesignated feature from data associated with the parameter signal. The predesignated feature is then aggregated by the control configuration and machine learning is applied to determine a degradation level of the function associated with the predesignated feature.

What is claimed is: 
     
       1. An elevator system comprising:
 a component adapted to perform a function; 
 a sensor configured to detect an operating parameter associated with the function; 
 a control configuration configured to receive a parameter signal from the sensor; extract a predesignated feature from data associated with the parameter signal, aggregate the predesignated feature, and apply machine learning to determine a degradation level of the function associated with the predesignated feature; and 
 a car adapted to travel in a hoistway, wherein the component includes a door assembly adapted to open and close for user access into and out of the car from and to a plurality of landings, wherein the door assembly includes a plurality of landing doors and the function is opening and closing of the plurality of landing doors, wherein the sensor is one of a plurality of sensors with each sensor located at a respective landing door of the plurality of landing doors. 
 
     
     
       2. The elevator system set forth in  claim 1 , wherein the feature includes vibration. 
     
     
       3. An elevator system comprising:
 a component adapted to perform a function; 
 a sensor configured to detect an operating parameter associated with the function; 
 a control configuration configured to receive a parameter signal from the sensor; extract a predesignated feature from data associated with the parameter signal, aggregate the predesignated feature, and apply machine learning to determine a degradation level of the function associated with the predesignated feature; 
 a feature generation module executed by the control configuration for extracting the predesignated feature from the parameter signal; 
 a fault detection module executed by the control configuration to analyze the predesignated feature and extract feature derivations from the predesignated feature indicative of abnormal operation; 
 a fault classification module executed by the control configuration to classify the feature derivations into respective fault classes; 
 a degradation estimation module executed by the control configuration to establish a learned degradation model; 
 wherein the control configuration includes a local controller and a server, and the local controller is configured to execute the feature generation module and the server is configured to execute the fault classification module and the degradation estimation module; 
 wherein the server is cloud-based; and 
 wherein the control configuration includes:
 a sensor hub configured to receive the parameter signal; 
 a mobile device configured to receive the parameter signal from the local controller, execute the feature generation module, execute the fault detection module, execute the fault classification module, and execute the degradation estimation module; and 
 the cloud-based server configured to communicate with the mobile device and store the learned degradation model for use by the degradation estimation module.